This table displays the selected game’s current values on four BGG outcomes (UsersRated, Average, GeekRating, Weight) along with my predictive model(s) estimated values for where these games are likely to end up.
Published | ID | Name | Type | UserRatings | Average | GeekRating | Weight |
2022 | Current | 259 | 7.04 | 5.69 | 1.70 | ||
Estimated | 1,800 | 7.54 | 6.52 | 1.88 |
To see more information about the game on boardgamegeek, click on the game’s ID or Name to go straight to the game’s profile page.
Description |
In Namiji, you are fishers from the Japan of yesteryear, navigating south of the Japanese archipelago, a few kilometers from the famous Tokaido road. You will need to have a fruitful day at sea to win the game. |
Information on the game’s designers, artists, families, mechanics and categories from BGG.
Published | ID | Name | Publisher(s) | Designer(s) | Artist(s) | Categories | Families | Mechanics |
2022 | Funforge | Antoine Bauza | Xavier Gueniffey Durin | Travel | Sports: Fishing / Angling | Point to Point Movement | ||
Country: Japan | Set Collection | |||||||
Crowdfunding: Kickstarter | Time Track | |||||||
Variable Player Powers | ||||||||
For the full profile of the selected game, click on the links in the table or go to https://boardgamegeek.com/boardgame/275215
For each of the predictions above, I used models that were trained on historical boardgamegeek data in order to predict the selected game. How did the model(s) arrive at their predictions? The following plot displays Shapley values to indicate what features were most influential for the model’s predictions. Anything in blue increased the model’s predictions, anything in red decreased the model’s predictions.
What games are similar to Namiji? I compare the selected game to every other game on boardgamegeek based on its mechanics, playing time, complexity, and game type. The tables below display the games which had the most similar data to Namiji. The reported similarity is the cosine similarity between the selected game and all other games published before 2021.
This similarity score ranges from -1 to 1, where similarity of 1 indicates that games are exactly identical (a game will have a score of 1 with itself).
Note: This analysis does not look at a game’s ratings on BGG, publisher, artist, or designer - it focuses on finding similar games based on how they play, not how they are rated or by their theme.
The table below displays games with the most similar data to Namiji. Click on a game’s ID or Name to go straight to its profile on boardgamegeek.
Rank | Published | ID | Name | Similarity | GeekRating | Average | Weight | UserRatings |
-- | 2022 | 1.00 | 5.69 | 7.04 | 1.88 | 259 | ||
1 | 2008 | 0.90 | 6.19 | 6.42 | 1.65 | 6,016 | ||
2 | 2011 | 0.89 | 5.48 | 4.73 | 2.00 | 48 | ||
3 | 2016 | 0.89 | 5.59 | 6.42 | 1.67 | 200 | ||
4 | 2017 | 0.88 | 6.08 | 7.02 | 2.07 | 1,178 | ||
5 | 1987 | 0.88 | 5.49 | 5.28 | 1.29 | 67 | ||
6 | 2012 | 0.87 | 5.98 | 6.66 | 2.20 | 1,246 | ||
7 | 2018 | 0.87 | 5.83 | 6.97 | 1.45 | 529 | ||
8 | 2012 | 0.86 | 5.54 | 6.02 | 1.89 | 168 | ||
9 | 1981 | 0.85 | 5.48 | 4.64 | 1.33 | 36 | ||
10 | 2015 | 0.85 | 5.52 | 6.53 | 2.43 | 45 | ||
11 | 2017 | 0.85 | 5.86 | 6.21 | 1.53 | 2,038 | ||
12 | 2019 | 0.83 | 5.53 | 6.84 | 2.00 | 45 | ||
13 | 2016 | 0.83 | 5.50 | 5.60 | 1.67 | 54 | ||
14 | 2019 | 0.83 | 5.49 | 5.21 | 1.00 | 30 | ||
15 | 2017 | 0.82 | 5.50 | 5.72 | 1.50 | 56 | ||
16 | 2004 | 0.82 | 5.70 | 6.15 | 1.99 | 862 | ||
17 | 1935 | 0.82 | 5.38 | 4.48 | 1.19 | 224 | ||
18 | 1954 | 0.82 | 5.53 | 6.67 | 1.00 | 48 | ||
19 | 2015 | 0.81 | 5.62 | 6.49 | 1.17 | 250 | ||
20 | 1978 | 0.80 | 5.52 | 6.18 | 1.90 | 72 | ||
21 | 2005 | 0.80 | 5.97 | 6.50 | 2.44 | 1,584 | ||
22 | 1978 | 0.80 | 5.51 | 5.95 | 1.17 | 73 | ||
23 | 1993 | 0.80 | 5.48 | 5.36 | 1.14 | 105 | ||
24 | 2014 | 0.79 | 5.85 | 7.28 | 2.36 | 460 | ||
25 | 1975 | 0.79 | 5.43 | 4.41 | 1.50 | 109 |
Click on the ID or Name for any game to go its profile page on boardgamegeek.
What games are similar to Namiji while also being highly rated by the boardgamegeek community? In this table I multiply every neighboring game’s similarity score by its GeekRating. The resulting score is an attempt to balance these two quantities.
Rank | Published | ID | Name | Score | Similarity | GeekRating | Average | Weight | UserRatings |
-- | 2022 | 5.69 | 1.00 | 5.69 | 7.04 | 1.88 | 259 | ||
1 | 2008 | 5.64 | 0.75 | 7.48 | 7.58 | 2.41 | 111,644 | ||
2 | 2008 | 5.60 | 0.90 | 6.19 | 6.42 | 1.65 | 6,016 | ||
3 | 2017 | 5.38 | 0.88 | 6.08 | 7.02 | 2.07 | 1,178 | ||
4 | 2012 | 5.21 | 0.87 | 5.98 | 6.66 | 2.20 | 1,246 | ||
5 | 2016 | 5.17 | 0.74 | 7.01 | 7.40 | 2.17 | 8,929 | ||
6 | 2018 | 5.07 | 0.87 | 5.83 | 6.97 | 1.45 | 529 | ||
7 | 2016 | 4.98 | 0.89 | 5.59 | 6.42 | 1.67 | 200 | ||
8 | 2017 | 4.96 | 0.85 | 5.86 | 6.21 | 1.53 | 2,038 | ||
9 | 2019 | 4.93 | 0.79 | 6.23 | 7.43 | 2.35 | 1,078 | ||
10 | 2011 | 4.88 | 0.89 | 5.48 | 4.73 | 2.00 | 48 | ||
11 | 1987 | 4.81 | 0.88 | 5.49 | 5.28 | 1.29 | 67 | ||
12 | 2005 | 4.78 | 0.80 | 5.97 | 6.50 | 2.44 | 1,584 | ||
13 | 2012 | 4.75 | 0.86 | 5.54 | 6.02 | 1.89 | 168 | ||
14 | 2016 | 4.70 | 0.79 | 5.97 | 6.57 | 1.94 | 1,423 | ||
15 | 2004 | 4.69 | 0.82 | 5.70 | 6.15 | 1.99 | 862 | ||
16 | 2015 | 4.68 | 0.85 | 5.52 | 6.53 | 2.43 | 45 | ||
17 | 1981 | 4.65 | 0.85 | 5.48 | 4.64 | 1.33 | 36 | ||
18 | 2014 | 4.65 | 0.79 | 5.85 | 7.28 | 2.36 | 460 | ||
19 | 2007 | 4.63 | 0.76 | 6.11 | 6.75 | 1.96 | 1,649 | ||
20 | 2019 | 4.59 | 0.83 | 5.53 | 6.84 | 2.00 | 45 | ||
21 | 2015 | 4.58 | 0.81 | 5.62 | 6.49 | 1.17 | 250 | ||
22 | 2012 | 4.56 | 0.76 | 6.02 | 6.71 | 2.02 | 1,330 | ||
23 | 2016 | 4.56 | 0.83 | 5.50 | 5.60 | 1.67 | 54 | ||
24 | 2019 | 4.54 | 0.83 | 5.49 | 5.21 | 1.00 | 30 | ||
25 | 2017 | 4.53 | 0.82 | 5.50 | 5.72 | 1.50 | 56 |
I first ran a dimension reduction method (PCA) to identify the main points of variation for games on BGG. The first 25 components explained roughly 50% of the overall variation in the dataset, so I computed the cosine similarity between every game’s score on these 25 components.
The following plot shows every game on BGG plotted on the first two principal components of variation. I then highlight where the selected game and its nearest neighbors fall on these first two components.
Similarity scores are based on 25 principal components - the next table displays the selected game and its neighbors profiles on each of these components.
What are these components? The following plot shows which features contribute the most to each component.